The Build‑Or‑Fake Prioritization Canvas: A 1‑Page Workflow to Decide Which Micro‑MVPs to Ship First

The Build‑Or‑Fake canvas is intentionally minimal: four axes you score 1–5 for each micro‑MVP idea, plus a short go/no‑go rule. The axes are: Development Cost (how many dev hours / dollars to produce a usable test), Learn Speed (how quickly you can run and collect meaningful results), Signal Quality (how reliable the data is as a proxy for real user behavior), and Go/No‑Go Criteria (the measurable threshold that will trigger building, iterating, or killing the idea).

Use the canvas as a decision amplifier, not a replacement for judgment. Sum the scores to get a raw prioritization, then apply strategic filters (market fit hypothesis, legal/compliance constraints, and team capacity). The goal is to expose assumptions and pick tests that deliver maximum actionable insight per unit of effort.

Avoid false precision by using concrete anchors for each 1–5 score. For Development Cost, anchor 1 = ‘landing page + copy (0–8 hours)’, 3 = ‘single‑page app + light backend (2–5 dev days)’, 5 = ‘full feature with integrations (3+ weeks)’. For Learn Speed, anchor 1 = ‘multi‑week cohort required’, 5 = ‘minutes to hours — click or signup signal’.

Signal Quality needs explicit definition: a click or CTA on a homepage is lower‑quality than an actual paid deposit, but much faster. Anchor Signal Quality from 1 = ‘soft interest (reads a blog post)’ to 5 = ‘monetary commitment or live usage’ and treat Confidence (evidence quality) separately if you want a fifth axis.

Below are three tested micro‑MVP patterns that map cleanly into the canvas. Each pattern includes a short build checklist, expected signals, and a practical go/no‑go rule you can copy. These are intentionally simple so you can run them fast and iterate from real user behavior.

Choose patterns that test the riskiest assumption for your business (willingness to pay, core retention loop, or distribution). Run the highest Weighted Learning Efficiency items first, then graduate promising fake experiments to higher‑commitment tests (e.g., fake door → pre‑order → single‑feature live).

Week 0: Fill the canvas for 5–8 candidate micro‑MVPs. Score each axis, compute Weighted Learning Efficiency, and pick 2–3 experiments that use different signals/channels so they don’t compete for the same users. Reserve one developer or maker to run experiments and one person to own analysis and follow‑ups.

Weeks 1–2: Execute fast. Examples: launch a paid pricing landing page (fake door) with a small ad spend or targeted email to an existing list; run a concierge workflow where you manually complete tasks while tracking time and conversion; ship a single‑feature app with telemetry for a small, invited cohort. After two weeks, compare results against the pre‑defined go/no‑go criteria and either build the full feature, iterate the experiment, or kill it and move on.

A successful fake‑door or concierge experiment should create a clear escalation path: (1) confirm signal with a higher‑commitment test (deposit, pre‑order, paid pilot), (2) implement a single‑feature live MVP that automates the most time‑consuming manual steps, and (3) instrument retention and monetization metrics before expanding scope. Always keep the original canvas entry and update its scores as evidence accumulates.

Treat prioritization as an ongoing loop rather than a one‑time gating decision. Use the canvas to re‑score opportunities after each experiment so the next round of priorities reflects actual learning, not original guesses.